# CloudResearch 1 - Political Knowledge ----------------------------------------
library(TAM)

A <- designMatrices(resp=cr1_political_knowledge)$A
A1 <- A
dimnames(A1)[[1]] <- colnames(cr1_political_knowledge)
A1[1,,] = A1[15,,]
A1[2,,] = A1[16,,]
A1[3,,] = A1[17,,]
A1[4,,] = A1[18,,]
A1[5,,] = A1[19,,]
A1[6,,] = A1[20,,]
A1[7,,] = A1[21,,]
A1[8,,] = A1[22,,]
A1[9,,] = A1[23,,]
A1[10,,] = A1[24,,]
A1[11,,] = A1[25,,]
A1[12,,] = A1[26,,]
A1[13,,] = A1[27,,]
A1[14,,] = A1[28,,]

cr1_pkscale_m <- tam.mml.2pl(cr1_political_knowledge,
                    group = cr1_language_group,
                    irtmodel = "2PL",
                    beta.fixed = cbind(c(1:3), 
                                       1, c(0, 0, 0)),
                    A = A1)

cr1_pkscale_m <- tam.mml.2pl(cr1_political_knowledge,
                    group = cr1_language_group,
                    irtmodel = "2PL",
                    beta.fixed = cbind(c(1:3), 
                                       1, c(0, 0, 0)),
                    A = A1,
          xsi.fixed = cbind(which(cr1_pkscale_m$xsi.fixed.estimated[,2] == 0), 
          which(cr1_pkscale_m$xsi.fixed.estimated[,2] != 0)))

cr1_pkscale_m2 <- tam.mml.2pl(cr1_political_knowledge,
                      group = cr1_language_group,
                      irtmodel = "2PL",
                      beta.fixed = cbind( c(1:3), 
                                          1, c(0, 0, 0)))

row.names(cr1_pkscale_m2$xsi) <- c(paste("polknow", "_e_", 1:14, sep = ""), 
                           paste("polknow", "_s_", 1:14, sep = ""))

cr1_pkscale_lr <- IRT.compareModels(cr1_pkscale_m,
                           cr1_pkscale_m2)

# CloudResearch 1 - Panethnic Identity -----------------------------------------
A <- designMatrices(resp=cr1_panethnic)$A
A1 <- A
dimnames(A1)[[1]] <- colnames(cr1_panethnic)
A1[1,,] = A1[6,,]
A1[2,,] = A1[7,,]
A1[3,,] = A1[8,,]
A1[4,,] = A1[9,,]
A1[5,,] = A1[10,,]

cr1_panethnic_m <- tam.mml.2pl(cr1_panethnic,
                     group = cr1_language_group,
                     irtmodel = "GPCM",
                     beta.fixed = cbind(c(1:3), 
                                        1, c(0, 0, 0)),
                     A = A1)

cr1_panethnic_m <- tam.mml.2pl(cr1_panethnic,
                     group = cr1_language_group,
                     irtmodel = "GPCM",
                     beta.fixed = cbind(c(1:3), 
                                        1, c(0, 0, 0)),
                     A = A1,
                     xsi.fixed = cbind(which(cr1_panethnic_m$xsi.fixed.estimated[,2] == 0), 
                                       which(cr1_panethnic_m$xsi.fixed.estimated[,2] != 0)))

cr1_panethnic_m2 <- tam.mml.2pl(cr1_panethnic,
                      group = cr1_language_group,
                      irtmodel = "GPCM",
                      beta.fixed = cbind( c(1:3), 
                                          1, c(0, 0, 0)))

cr1_panethnic_lr <- IRT.compareModels(cr1_panethnic_m,
                            cr1_panethnic_m2)

# CloudResearch 1 - Efficacy Beliefs -----------------------------------------------
A <- designMatrices(resp=cr1_efficacy)$A
A1 <- A
dimnames(A1)[[1]] <- colnames(cr1_efficacy)
A1[1,,] = A1[7,,]
A1[2,,] = A1[8,,]
A1[3,,] = A1[9,,]
A1[4,,] = A1[10,,]
A1[5,,] = A1[11,,]
A1[6,,] = A1[12,,]

cr1_efficacy_m <- tam.mml.2pl(cr1_efficacy,
                        group = cr1_language_group,
                        irtmodel = "GPCM",
                        beta.fixed = cbind(c(1:3), 
                                           1, c(0, 0, 0)),
                        A = A1)

cr1_efficacy_m <- tam.mml.2pl(cr1_efficacy,
                        group = cr1_language_group,
                        irtmodel = "GPCM",
                        beta.fixed = cbind(c(1:3), 
                                           1, c(0, 0, 0)),
                        A = A1,
                        xsi.fixed = cbind(which(cr1_efficacy_m$xsi.fixed.estimated[,2] == 0), 
                                          which(cr1_efficacy_m$xsi.fixed.estimated[,2] != 0)))

cr1_efficacy_m2 <- tam.mml.2pl(cr1_efficacy,
                         group = cr1_language_group,
                         irtmodel = "GPCM",
                         beta.fixed = cbind( c(1:3), 
                                             1, c(0, 0, 0)))

cr1_efficacy_lr <- IRT.compareModels(cr1_efficacy_m,
                               cr1_efficacy_m2)

# CloudResearch 1 - Minority Representation -----------------------------------------------
A <- designMatrices(resp=cr1_minrep)$A
A1 <- A
dimnames(A1)[[1]] <- colnames(cr1_minrep)
A1[1,,] = A1[4,,]
A1[2,,] = A1[5,,]
A1[3,,] = A1[6,,]

cr1_minrep_m <- tam.mml.2pl(cr1_minrep,
                     group = cr1_language_group,
                     irtmodel = "GPCM",
                     beta.fixed = cbind(c(1:3), 
                                        1, c(0, 0, 0)),
                     A = A1)

cr1_minrep_m <- tam.mml.2pl(cr1_minrep,
                     group = cr1_language_group,
                     irtmodel = "GPCM",
                     beta.fixed = cbind(c(1:3), 
                                        1, c(0, 0, 0)),
                     A = A1,
                     xsi.fixed = cbind(which(cr1_minrep_m$xsi.fixed.estimated[,2] == 0), 
                                       which(cr1_minrep_m$xsi.fixed.estimated[,2] != 0)))

cr1_minrep_m2 <- tam.mml.2pl(cr1_minrep,
                      group = cr1_language_group,
                      irtmodel = "GPCM",
                      beta.fixed = cbind( c(1:3), 
                                          1, c(0, 0, 0)))

cr1_minrep_lr <- IRT.compareModels(cr1_minrep_m,
                                   cr1_minrep_m2)



